Great Race to the Bottom

Member Group : Keith Naughton

Normally a Presidential candidate viewed unfavorably by 60% of voters would not stand a chance. But Donald Trump gets to run against Hillary Clinton and her own mind-numbing unpopularity. The 2016 contest for President is not so much a race, but a mutual free fall. Who wins? A dyspeptic Alf Landon? Or a corrupt George McGovern?

While both Trump and Clinton had their ups and downs within their own nomination contests, the polling differences between them and the public perception of them has changed little. For a brief period of time it looked as though there just might be a real shift in the race toward Trump. In the aftermath of the release of the FBI report on Clinton’s e-mails Trump went on a streak of 6 straight polls where he led Clinton. Prior to that burst of success, Trump only led Clinton three times out of over 50 polls since January 1st. In the aftermath of the GOP convention, Trump reached a high water mark with a 47-40 lead in the LA Times/USC poll.

Since then, the race has reverted to the status before the conventions, Clinton with a small, but consistent lead and both candidates still incredibly unpopular. On the last day of the Democratic convention Rasmussen (which was the one poll that consistently put Trump in the lead – if only by one point) had Clinton ahead 43-42. A series of polls from Reuters, PPP and CBS have put Clinton ahead by an average of 5 points.
The bottom line is that this race is unlikely to change without a major gaffe by either candidate or an extraordinary outside event. Unlike past races where undecided voters slowly made clear moves toward candidates, the electorate in 2016 has shown itself locked into consistent pattern that has weathered every Trump tweetstorm and Clinton foot-in-mouth moment.

To see how consistent this race has been polling data from January 1st by pollsters graded at least a B+ from the site FiveThirtyEight and was drawn from the RealClearPolitics archives, was examined. Polls during the conventions were not included. June 7th, the date Clinton clinched the Democratic nomination, is used as a dividing point.

By far the most salient feature of this election is the staggering unpopularity of both major party nominees. It has been over a year since more Americans viewed Hillary Clinton favorably than not (+7 on July 12, 2015). Clinton did crawl back to -1 in a May 1st CNN poll, but that was sandwiched between ratings of -12, -15, -17 and -19, making that close call a bit dubious. Overall, Clinton has averaged a favorable/unfavorable split of 39/55 this year. And her overall ratings are trending down. Her favorable level has edged up slightly, from 38.9% before June 7th to 39.4%, but her unfavorable level has increased more, from 54.4% to 55.2%.

Donald Trump has never been net favorable. The closest yet was -5 in CNN’s first post-GOP convention poll (sandwiched between polls of -19 and -26). Before that, Trump’s "high water" mark was -14 on September 15, 2015. With the lone exception of one CNN poll, Trump’s favorable level has never exceeded 43% and his unfavorable has never dropped below 55%. The good news is that his standing has improved. Trump has gone from a favorable level of 31.6% and unfavorable of 61.2% before June 7th to a split of 34.5% favorable and 59.7% unfavorable after.

Unfortunately for Trump and Clinton the public seems to have made up its mind. Not only have Trump and Clinton been consistently unpopular, there has been limited change over time. For both candidates, the standard deviation for their unfavorable numbers has been low and declining. (Note: Standard deviation is a measure of the spread in a set of numbers. In the case of polling numbers, a high number indicates voters are changing their minds. A low number means minds are not changing.) For both Trump and Clinton, the standard deviation of their unfavorable level has fallen.
Trump has gone from 3.78 (ranging from 57.5% to 65%) to 3.37 (narrowing to 56.3% to 63%). Clinton-haters are even more settled in, with her standard deviation falling from 3.11 (ranging from 51.3% to 57.5%) to 2.27 (narrowing to 57.4% to 62%).

At the same time, while certainty over who voters dislike has gone up, certainty over who they like has gone down. The standard deviation for Trump’s favorable rating has increased from 4.47 to 5.09. Clinton’s favorable standard deviation has also increased from 4.68 to 4.95. Increasingly, voters know who they don’t like, but are less able to settle on who they like.

The conventions did little for either candidate. In the three polls since the end of the Democrats’ infomercial, Clinton averaged a 41/53 split and Trump averaged a 37/56 split – slight improvements, but hardly game-changing.

These dismal favorability numbers help explain a curious feature of the ballot test: things got worse for both candidates after they clinched. Prior to the GOP convention, both Trump and Clinton saw their ballot test numbers go down after clinching their respective nominations. Trump dropped just over one point, from 40.9% to 39.7%. Clinton had a much greater fall, dropping from 48% to 44.2%. The narrowing of the race has been more due to Clinton sinking than Trump rising.

In the three polls since the Democratic convention, the average ballot test has seen a Clinton lead of 45.0% to 40.7% – a bounce of merely one point for Trump and less than one point for Clinton. Just over 32 million Americans watched Trump’s acceptance speech and 29.8 million watched Clinton – and the result was a big yawn.

The way the 2016 race is developing is completely at odds with how Presidential races traditionally unfold, where a large undecided group moves towards each candidate as the race develops. Getting to the point where your favorability is greater than your opponent’s – or, more importantly, your unfavorability is less – is a critical goal.

In 2012, Obama’s favorable rating remained practically constant: 50.7% through the GOP primary contest, dropping slightly over the summer to 50.3%, then edging up in the fall to an average of 50.9%. The standard deviation also fell over time, from 4.71 during the primaries down to just 2.98 in the heat of the general election. Obama’s unfavorable also stayed fairly constant, only rising from 44.4% to 45.3%. The standard deviation fell from 2.82 to only 1.82 – mirroring the 2016 contest.

Typical for a challenger, Romney’s favorable rose over time, from 37.3% in the primaries to 47.6% in the general election, while his unfavorable stayed fairly constant, only increasing from 43.5% to 45.3%. Not surprisingly, Romney has a favorable high standard deviation during the primaries of 6.53, falling to 3.74 in the general; consistent with a public becoming more familiar with a new candidate.

For the ballot test, Obama was very consistent, rising from 47.6% in the primaries to 48.3% in the general. Romney rose from 42.3% to 46.2%. The final weekend of polls showed a separation of just 1%, with neither polling over 50%. Obama’s eventual victory (51% to 47.2%) was a result of many factors like all close races, but the bottom line is that Romney could not push Obama under 50% approval.

What a difference four years makes. In a race where Clinton fulfills the role of the incumbent with Trump as the challenger, the poor polling by Clinton should advantage Trump and would certainly have made any of the other GOP hopefuls a strong favorite (with the possible exception of Ted Cruz). However, Trump’s numbers have been even worse than Clinton’s. Trump won the GOP nomination fair and square, but his performance in the primaries was weaker than any previous Republican nominee. Clinton is in no position to look down on Trump as she was pushed to the limit by a second-tier protest candidate who threw away his best issue at the start of the campaign.

This election is shaping up to be a difficult war of attrition where the decision will be made by voters who will be toting up a list of reasons not to vote for Trump or Clinton. Given the consistent lack of movement, it may well be that any significant change will only result of events beyond the control of the candidates or a colossal blunder. For two candidates prone to lapses in judgment, figuring out which candidate, Trump or Clinton, will anger voters more is an even-money bet.